Prognosis of Lithium-Ion Batteries Considering Cycle and Storage Conditions

Abstract

Lithium-ion batteries are used as a major power source for electric vehicles (EV) due to their high energy density. When the battery is repeatedly charged and discharged during the operation of the vehicle, the capacity is gradually deteriorated, which is divided into cycle degradation and storage degradation. In order to predict the accurate life of the battery, combination of the two degradations must be considered reflecting the actual usage conditions. However, composite (i.e., both cycle and storage conditions) tests take relatively longer than the tests for each conditions. In this study, we proposed a method for predicting the remaining useful life (RUL) of a complex situation using the results of respective tests on cycle and storage conditions, which require relatively short test times. Based on the results of respective tests, the model-based approach using particle filter is applied to predict RUL of a lithium-ion battery, which was derived as a probability distribution. By modeling the joint probability distribution using copula, RUL of composite condition considering both cycle and storage degradations. It was verified by comparison with the distribution of the RUL derived using the composite tests results.